Mikko-Jussi Laakso

Generative Artificial Intelligence (GenAI) tools are permeating education but whether cognitive dispositions moderate their benefits remains unclear. In the present work, we examined whether Cognitive Reflection and Intolerance of Uncertainty moderate the relationship between prior GenAI experience and collaborative task effectiveness. Specifically, secondary ( n = 114) and tertiary ( n = 57) students, distributed across groups, completed a mystery‐solving task with unrestricted access to a GenAI tool of their choice. Person‐centred analysis identified no cognitive‐attitudinal profiles as the solution collapsed onto three groups defined by the Cognitive Reflection Test (CRT) score. In exploratory comparisons, the High‐CRT group reported greater post‐task prompting self‐efficacy than the Low‐CRT group, despite no detectable difference in performance. Cognitive reflection did not moderate the experience‐effectiveness link at the group level and uncertainty intolerance did not predict AI reliance behaviours. Therefore, cognitive dispositions appear to relate more closely to the subjective experience of working with AI than to objective task outcomes.

Despite the documented benefits that Augmented Reality (AR) offers for language education, teacher adoption remains constrained by barriers that differ fundamentally in tractability. For instance, infrastructure limitations require systemic investment whereas attitudinal concerns may respond to awareness-raising and professional development. The present study examines the barrier hierarchy among Greek English as a Foreign Language (EFL) teachers and the relationship between AR knowledge and barrier perceptions. Using a mixed-methods design, EFL teachers ( N = 266) completed a survey with subsets participating in focus groups ( n = 35) and semi-structured interviews ( n = 18). Quantitative analysis revealed that infrastructure barriers exceeded attitudinal barriers. AR knowledge was associated with lower attitudinal barriers but showed no relationship with infrastructure perceptions, indicating that awareness interventions cannot overcome material constraints. Evidence of student learning gains emerged as the strongest condition for adoption readiness, outweighing even resource provision. The qualitative data broadly converged with the barrier hierarchy but also exposed heterogeneity—including divergent stances and context-specific reasoning that the quantitative aggregate obscured. Exploratory analyses further showed that barrier profiles varied by teaching level. It is, therefore, concluded that, where infrastructure constraints dominate, awareness-raising alone may be insufficient to accelerate AR adoption and that coordinated investment across infrastructure, knowledge, and pedagogy offers a more plausible path.

This study explores students' emotions in the undergraduate chemistry laboratory, focusing particularly on emotional responses to making mistakes. Using an ecological momentary assessment design, we collected on‐task data from 240 students and 1635 learning situations during the laboratory sessions. A multilevel structural equation modeling framework, including a model with random slopes, was used to examine how situational factors (i.e., type of mistake and support in solving the mistake) and individual factor (i.e., gender, performance level, and subjective task values) affected students' emotional responses. Our findings suggest that mistakes typically trigger negative activating emotions in students. Individual factors also influenced emotional responses—well‐performing students were less prone to elevated anxiety and frustration, remaining more relaxed in mistake situations, whereas students with high perceived emotional costs of laboratory activities experienced a more intense increase in anxiety in mistake situations. The type of mistake played an important role: careless mistakes were associated with more positive emotions, whereas students experienced increased anxiety and frustration and decreased positive emotions when the mistake stemmed from skill or knowledge gaps. The form of support also mattered: teacher support was linked to intensified anxiety and frustration in students, while peer support showed no statistically significant impacts on emotional levels. To foster meaningful learning, targeted support in emotional regulation should be provided for lower‐performing students and those with high perceived emotional costs in mistake situations, particularly when mistakes stem from a lack of skill or knowledge.

Augmented Reality (AR) has been found to produce significant effects on individual learning outcomes but its impact on collaborative applications remains moderate. Existing AR frameworks emphasize individual instructional design, whereas frameworks for collaborative learning rarely engage with the spatial and device-mediated affordances of mobile AR. In response to this inadequacy in the literature, we introduce the Mobile Augmented-Reality Storytelling for Vocational Education and Training (MARS-VET) framework, a four-dimensional conceptual architecture that integrates Computer-Supported Collaborative Learning (CSCL) scripting principles with mobile AR affordances for collaborative English as a Foreign Language (EFL) writing in Vocational Education and Training (VET) settings. MARS-VET synthesizes theoretical perspectives across four dimensions: contextual anchoring, which embeds activities within authentic workplace scenarios; collaborative orchestration, which structures group interaction through macro- and micro-level scripts; competency cultivation, which sequences writing progression from model-based reproduction toward autonomous professional text production; and capacity building, which addresses the professional-development requirements of implementing educators. Content validity was established through expert panel evaluation involving international specialists (N = 11) who rated the framework against 36 items using a four-point relevance scale and provided additional qualitative feedback. The Scale-level Content Validity Index (S-CVI/Ave = 0.91) exceeded established thresholds, with all four dimensions achieving satisfactory item-level indices. Experts reached unanimous agreement on items addressing workplace scenario identification and co-located access to linguistic resources. Qualitative feedback led to terminology refinements and clarification of orchestration mechanisms. The framework offers VET institutions and educators a reference for the design and evaluation of collaborative AR experiences in an area where integrative frameworks have so far been lacking.

This paper analyses computational models of Knowledge Tracing (KT), which address the complex sequence-modelling task of predicting dynamic, unobservable latent user states from historical interaction logs. First, we propose a comprehensive taxonomy identifying nine distinct and interconnected KT model families: psychometric; Bayesian; machine learning; deep learning; graph-based; temporal/sequential; multi-task; contrastive/self-supervised; and domain-adaptive. Secondly, we trace the historical evolution of KT architectures, from the foundational psychometric methods of the 1950s to the modern integration of attention mechanisms and graph neural networks. Thirdly, we systematically evaluate nine lightweight representative computational models—one from each category—across two large-scale datasets: ASSISTments 09-10 and DigiArvi 2025. We measure predictive calibration using accuracy, F1 score, ROC-AUC, average precision, and log loss under a strict computational time budget. Finally, our rigorous empirical analysis demonstrates that multi-task and temporal/sequential architectures yield the highest performance. Specifically, Fine-Grained Knowledge Tracing (FKT) achieved the best results on the DigiArvi dataset (accuracy: 0.77; F1 score: 0.85), while Temporal Item Response Theory (TIRT) performed best on the ASSISTments dataset (accuracy: 0.70; F1 score: 0.75). Traditional baselines, such as Logistic Regression (LR), remain highly competitive. Consequently, we advocate a shift towards 'Green AI' and standardized benchmarking to address the field’s fragmented evaluation standards, as we identify diminishing returns from increasing model complexity. Future research must leverage generative Artificial Intelligence (AI) and causal inference to move beyond simple prediction toward agentic AI systems capable of active pedagogical intervention.

Computational Thinking (CT) is widely recognised as a transversal competence essential for learning, problem solving, and knowledge transfer across disciplines. However, its effective integration into school education remains strongly dependent on the availability of assessment instruments that are pedagogically meaningful, psychometrically sound, and applicable across diverse educational contexts. This paper presents COMATH, a cross-national assessment instrument designed to evaluate CT in students aged 9–14. The instrument adopts a phase-based development and validation framework that integrates Bebras-inspired tasks, Item Response Theory, factor-analytic methods, learning analytics, and teacher and student feedback. The assessment was iteratively developed and piloted between 2023 and 2025 in six European countries, with data collected from 6,480 students and 155 teachers. The findings demonstrate that a phased assessment approach enables systematic calibration of task difficulty, robust evaluation of item functioning, and meaningful interpretation of student performance across age groups and national contexts. The results further highlight how well-designed CT assessment can support instructional decision-making rather than serve solely as a summative measure. The study argues for conceptualising CT assessment as a dynamic and iterative process that links measurement, psychometric validation, and pedagogical use in school education.

Elokuussa 2021 Suomen Opetus- ja kulttuuriministeriö (OKM) käynnisti kaksivuotisen esiopetuksen kokeilun. Kokeilussa osa kahden ikäkohortin (vuonna 2016 ja 2017 syntyneistä) lapsista osallistui esiopetukseen kokeiluryhmissä kahden vuoden ajan. Verrokkiryhmien lapset osallistuivat esipetukseen totuttuun tapaan yhden vuoden ajan. Kokeilun tavoitteena oli lisätä koulutuksellista tasa-arvoa, ja selvittää muun muassa kaksivuotisen esiopetuksen vaikutuksia lasten kehitys- ja oppimisedellytyksiin, sosiaalisiin taitoihin ja itsetunnon muodostumiseen. Kokeilun vaikutusten arviointi toteutettiin lasten ensimmäisen luokan syksyllä. Arviointi koostui oppilaiden suorittamista tehtävistä ja opettajien tekemistä oppilaskohtaisista arvioista. Oppilasarvioinnin kokonaisuudesta toimitettiin luokkien opettajille ja koulujen rehtoreille digitaalinen palaute molemmista kohorteista, joka sisälsi tietoa muun muassa oppilaiden luku- ja laskutaidosta oppilas- ja luokkatasoisesti sekä suhteessa koko ikäluokkaan. Opettajille ja rehtoreille haluttiin tarjota ajantasaista tietoa heidän oppilaistaan, mutta myös tukea heidän sitoutumistaan tutkimukseen. Tämä raportti koskee kaksivuotisen esiopetuskokeilun ensimmäisen luokan oppilasarviointien palautetta ja sen hyödyntämistä. Raportin tulokset perustuvat keväällä 2024 (I kohortti) sekä loppuvuodesta 2024 ja alkuvuodesta 2025(II kohortti) kerättyyn opettajien ja rehtoreiden kyselyaineistoon sekä keväällä 2024 kerättyyn opettajien ja rehtoreiden haastatteluaineistoon. Opettajien ja rehtoreiden kyselyissä selvitettiin alkuopetuksessa käytettäviä arviointimenetelmiä, vastaajien arvioita tutkimuksen keskiössä olevan palautteen vaikutuksesta heidän käsityksiinsä oppilaiden yleisestä suoriutumisesta, opettajien ja rehtoreiden käytäntöihin sekä opettajien ja rehtoreiden käsityksiin oppilaiden tuen tarpeista ja tuen tarpeista tehtävään päätöksentekoon. Lisäksi rehtoreiden kyselyssä kartoitettiin koostetiedon käyttöä koulujen tietojohtamisessa. Lisäksi kaikkien yllä kuvattujen muuttujien yhteyksiä opettajien ja rehtorien työkokemusvuosiin tarkasteltiin. Haastatteluissa samoja teemoja syvennettiin. Kyselyihin vastasi yhteensä 617 opettajaa ja 329 rehtoria. Yksilöhaastatteluihin osallistui 12 opettajaa ja 9 rehtoria. In August 2021 the Finnish Ministry of Education and Culture launched a trial for two-year pre-primary education. In the trial some children from two age cohorts (birth years 2016 and 2017) participated in pre-primary education for two years in trial groups. Children in control groups participated in pre-primary education for one year as usual. The aim of the trial was to increase educational equality but also examine how two-year pre-primary education effects children’s development and learning prerequisites, social skills and development of self-esteem. The effects of the trial were evaluated in the beginning of the first grade. Evaluation consisted of assessment tasks for students as well as student-specific assessments made by teachers. Students’ class teachers and school principals received digital feedback for both age cohorts of the trial about this assessment. Feedback included student-specific information for example about student’s literacy and mathematics skills both on class level and compared to the whole age cohort. The aim was to offer up-to-date information for teachers and principals about their students but also support their commitment to the study. This report is about feedback and use of feedback from first-grade student assessments in a two-year preprimary education trial. The results of the report are based on a survey of teachers and principals collected in the spring of 2024 and at the beginning of 2025, as well as interview data on teachers and principals collected in the spring of 2024. The surveys focused on assessment methods used by first-grade classroom teachers and participants’ evaluations of the effectiveness of feedback in several areas. These areas included how feedback affected teachers' and principals' perceptions of student performance, influenced their practices, and shaped their understanding of individual students’ special support needs and related decision-making. Also, the use of aggregate data in school knowledge management was surveyed. The links of all the variables mentioned above to teachers’ and principals’ work experience were also investigated. The same themes were delved deeper into during the interviews. Surveys were completed by 617 teachers and 329 principals. Twelve teachers and nine principals participated in individual interviews

Background: Previous research has shown that testing differs significantly from other classroom activities and is associated with heightened negative emotions and lower levels of positive emotions. However, relatively little is known about students' emotions surrounding testing, particularly in higher‐stakes assessment settings. Aims: This study aims to examine how students' levels of four emotions (i.e., happiness, relaxation, anxiety and boredom) develop from pre to‐post‐test, and it investigates how individual factors (i.e., gender, grade level, perceived mathematical competence and test performance), impact students' emotional states and moderate their emotional trajectories. Sample: The sample ( N = 2179) consists of 692 third‐grade, 605 sixth‐grade, 413 eighth‐grade and 469 ninth‐grade students from various schools across Finland, who participated in a digital, semi‐high‐stakes, end‐of‐year mathematics assessment. Methods: An in‐situ approach was used to assess students' emotions immediately before and after testing. Analyses were conducted using linear mixed‐effects modelling to account for the repeated‐measurements structure. Results and Conclusions: Students generally reported lower positive emotions after the assessment. The measured individual factors significantly predict both students' emotional states and their development during the assessment. Boys reported higher levels of positive emotions and lower anxiety, while younger students remained more positive during the assessment. Students who perceived themselves as competent experienced higher levels of positive and lower levels of negative emotions, whereas students who performed poorly showed a decline in positive emotions during the assessment. Future research could focus on whether support for emotional regulation affects student performance in test situations.

This study investigates the role of mistakes and affective experiences during online pre-lab activities in predicting students’ situational engagement (conceptualized here as a simultaneous experience of interest, skill, and challenge, i.e. optimal learning moments) in subsequent laboratory sessions in an undergraduate chemistry laboratory course ( n = 256). The data collection followed an ecological momentary assessment design. We specified multilevel structural equation models (MSEMs), including two- and three-level structural equation models, to examine how mistakes impacted students’ situational engagement during pre-lab activities and subsequent laboratory sessions. The findings indicate that mistakes in pre-lab tasks were associated with lower perceived skill and higher experience of challenge during that task, but did not predict students’ interest, skill, challenge, or situational engagement in the subsequent laboratory session. Autoregressive effects from pre-lab activities on students’ situational engagement during lab sessions were observed across all elements of engagement, while skill and challenge during pre-lab activities also predicted higher interest in the subsequent laboratory session. Based on our findings, we propose that while mistakes in the pre-lab activities do not play a significant role in predicting students’ engagement upon entering the laboratory, affective experiences during pre-lab activities can play a significant role in predicting students’ engagement in the laboratory, and that laboratory engagement could be enhanced by providing students with sufficiently challenging pre-lab activities.